Lenalidomide Administered for the Initial Treatment of Chronic Lymphocytic Leukemia (CLL) Patients- In Vivo Modulation of the Leukemia Cell Surface Phenotype and Association with Tumor Flare Reaction (TFR).

Abstract 3440

Poster Board III-328

Lenalidomide (L), an oral immunomodulatory agent used for the treatment of MM and MDS and has clinical activity in patients (pts) with relapsed/refractory CLL. When administered to CLL pts L is often associated with a TFR. TFR is painful enlargement of the lymph nodes and/or spleen. The mechanism of TFR is unknown. Previous reports have demonstrated upregulation of costimulatory molecules on leukemia cells with in-vitro L exposure. This led to the hypothesis that B cell activation is responsible for the TFR, possibly allowing for greater immune responses and infiltration of the lymph nodes/ spleen with inflammatory cells. Another hypothesis is that TFR results from the modulation of chemokine receptors allowing malignant lymphocytes to experience greater homing to the lymph nodes. We designed a phase II study to evaluate the efficacy and tolerability of L and rituximab as initial therapy of CLL. This previously untreated pt population provides a fairly homogenous population to assess the impact of in-vivo L treatment on CLL cells and test these hypotheses. To assess the impact of L on the CLL cells we designed a 3 week L run-in and performed ex-vivo analysis of the CLL cells prior to rituximab. All pts received L 2.5mg daily for the first 7 days and were escalated to 5 mg if they lacked significant toxicity on day 8. 20 pts have been enrolled on study and received at least 3 weeks of L. We report the early results of corollary science phenotyping the leukemia cells of 12 pts through the relative expression of a number of surface markers prior to and 21 days post initiation L. 9 of these pts experienced the TFR and 3 did not. 7 pts were on 5mg of L on day 21 and 5 on 2.5mg. Leukemia cells were stained for CD19, CD20 and chemokine receptors CXCR4, CXCR5, and CCR7 and for costimulatory molecules CD40, CD80, CD86, and CD54. Mean fluorescent intensity (MFI) was determined on CD19+ lymphocytes. MFI ratio (MFIR) was calculated by dividing the MFI for each stain by the isotype control of the respective sample. Two-sided student t test were used to assess statistical significance. Increased expression of CD40 was observed in 9/12 pts, (MFIR p 0.09) including 7/9 pts with TFR. Mean fold increase of CD40 MFIR for those with TFR was 1.13 compared with 1.26 in those without TFR (p = NS). CD86 increased in 6/12 including all 3 pts who did not experience TFR. Mean fold increase of CD86 MFIR for those with TFR was 1.11 and 1.6 in those without TFR (p = NS). CD80 MFIR increased slightly above pretreatment levels in 3/12 pts (3/9 with TFR). CD54 (ICAM) was evaluated in 9 pts and increased in 6. Median MFIR of CD54 following in vivo L increased from 14.7 to 24.8 (p = 0.066). Mean fold increase of CD54 MFIR for 2 pts without TFR was 2.08 compared with 1.82 for those with TFR (p= NS). MFIR for CXCR4 decreased in 10/12 pts during L including 8/9 pts with TFR. Median MFIR for CXCR4 prior to treatment was 67.9 and decreased to 37.5 (p = 0.024). Pts who experienced TFR had a more pronounced decrease in CXCR4 expression with MFIRs following L treatment mean 56% of pretreatment levels compared to the pts without TFR at 72% (p = NS). To assess if decreased CXCR4 expression could be the result of CLL cells mobilized from microenvironment niches we assessed the density of CXCR4 on 4 pts. All 4 pts had an increase in the population of CD19+ cells with dim CXCR4 expression from 10.2% prior to therapy to 26.4% following L (p = 0.058). MFIR for CXCR5 decreased in 9/12 pts (7/9 with TFR). For the 12 pts median CXCR5 MFIR 122.8 prior to treatment and 75.3 following 21 days of L (p = 0.045). Median MFIR for those who did or did not experience TFR was 79% and 80% of pretreatment levels respectively. There was not a significant change in CCR7 for the 12 pts or clear relationship to TFR. Previous studies reported in vitro downmodulation of CD20 on CLL cells exposed to L. As this clinical study is evaluating the combination of L and rituximab, we assessed CD20 levels on CLL cells exposed to L in vivo. Median MFIR for CD20 prior to treatment was 10.5 and was 8.5 following 21 days of L (p = 0.064). In the 12 pts evaluated CD20 expression increased in 5 pts (mean increase 24% (range 5-58%)) and decreased in the other 7 (mean decrease 31% (range 5-42%)). All pts clinically responded to the addition of rituximab with decrease in absolute lymphocyte count. These data show no clear relationship to the upregulation of markers of B cell activation on circulating CLL cells and the occurrence of TFR. There is early evidence of in-vivo down modulation of chemokine receptors CXCR4 and CXCR5, however this phenomenon is not markedly different in pts who have experienced TFR.